An Introduction to Applied Linguistics

participants to detect missing vowels, the researchers found that Arabic ESL 
learners were considerably slower and less accurate than non-Arabic counterparts. 
Ryan and Meara (1991) conclude that their findings confirm the earlier results 
that L1 orthography has a long and lasting impact on L2 processing.
In addition to the influences of L1 orthography, researchers have also investigated 
the influences of L1 phonology on L2 word recognition and, consequently, on L2 
reading. L1 learners show preferences for acquiring new vocabulary (in their L1) 
with phonological patterns (‘phonotactics’) that are already familiar to them or 
that are already in their repertoire. They tend to avoid or acquire less readily words 
with unfamiliar sound patterns (Gathercole and Baddeley, 1989). In L2 contexts, 
Feldman and Healy (1998) reported an interesting experiment designed to test 
whether L2 students might actually avoid learning the meanings of L2 words with 
phonotactic patterns unfamiliar to them from their L1. These authors studied a 
group of native speakers of Japanese at intermediate levels of ESL instruction and 
found that the learning of common, high-frequency English words was affected 
by the similarity or difference of the phonological patterns of those words from 
phonological patterns in Japanese. Meanings of common English words with 
familiar L1 phonotactic patterns were easier to acquire than the meanings of 
common English words with unfamiliar L1 phonotactic patterns.
Implications: Teachers need to be aware that L2 learners coming from an L1 with 
a different orthographic system may be disadvantaged – particularly at beginning 
reading levels – not only because they have to learn a new orthographic system, 
but because they may also need to develop new processing mechanisms more 
suitable to the L2.
Vocabulary
Issues in Vocabulary and L2 Reading Development
There are a number of issues that centre on the contributing role of vocabulary 
knowledge for L2 reading abilities:
• The number of words needed to read L2 texts independently and for instructional 
uses.
• The role of context in L2 vocabulary acquisition and in the guessing/guessability 
of word meaning in L2 reading.
• The role of dictionaries of various kinds and the use of cognates in L2 vocabulary 
acquisition and in L2 reading.
• The ways L2 learners go about the task of acquiring vocabulary in the L2.
• The role of extensive or pleasure reading in the ‘incidental’ acquisition of L2 
vocabulary and the role of vocabulary instruction.
• The impact of various kinds of vocabulary instruction on L2 vocabulary 
development.
The first three issues above are surveyed briefly in this section. (See Chapter 3, 
Vocabulary, for more on these issues and for discussion of the last three issues.)

---

221
Reading
How Much L2 Lexis is Needed?
Several researchers have addressed the issue of how much vocabulary is necessary 
for L2 reading, from different perspectives. Laufer (1989) addressed the question 
in terms of percentage of text-lexis necessary for comprehension of academic 
literature by native speakers of Hebrew and Arabic in a university EAP course. 
She found significant differences at the 95 per cent level of text coverage, and 
concluded that L2 readers had a significantly higher chance of being a ‘reader’ 
if they understood 95 per cent of the text’s word tokens. Nation (2006; Hu and 
Nation, 2000) found that the percentage necessary might be closer to 98 per cent. 
Nation (2006) addressed the question in terms of the vocabulary size needed to 
read short, unsimplified novels for pleasure. Their results showed that in order to 
achieve 98 per cent coverage of the running words in such texts, that a vocabulary 
size of about 8000–9000 word families would be needed (and 9000 word families 
is likely to translate into more than 35,000 individual word forms) (see Cobb, 
2009; Schmitt, Jiang & Grabe, in press).
The seminal study in this area was conducted by Hazenberg and Hulstijn 
(1996). In a very carefully designed and executed study with Dutch native 
speakers reading first-year university level materials, Hazenberg and Hulstijn 
(1996) first assessed the representativeness of more than 23,000 words (lemmas) 
taken from a dictionary to cover a 42 million-word corpus of contemporary 
written Dutch. They found that, with frequency as a criterion, text coverage 
substantially increased with up to 11,123 words but not beyond. Next, Hazenberg 
and Hulstijn (1996) assessed the representativeness of the same 23,000 words to 
cover first-year university reading materials. They found that the coverage of the 
academic corpus did not differ from the coverage of the larger general corpus. In 
the third part of the study, they developed and administered a vocabulary test 
aimed at measuring receptive knowledge of more than 18,000 content words 
of the 23,000 words. From these results they concluded that the minimal size 
vocabulary needed for university study is 10,000 base words, clearly a larger 
vocabulary size than required for reading everyday unsimplified texts such as 
newspapers or novels.
The major transforming study identifying the wider range of vocabulary 
needed for academic L2 reading is that of Nation (2006). He first argued 
persuasively that adequate text comprehension is typically achieved when a 
reader knows at least 98 per cent of the words appearing on a page. He then 
analysed word-family frequency lists from the British National Corpus and 
determined that good comprehension of written texts requires between 8000 
and 9000 word families (or somewhat more than 36,000 individual word types; 
see Schmitt, 2008).
The Role of Context in Guessing/Guessability of Word Meaning 
in L2 Reading
In a seminal L2 study, Bensoussan and Laufer (1984) investigated use of context 
by university-level EFL students in translating words into their native language. 
Through analysis of student answering patterns they determined that context 
helped lexical guessing in only 13 per cent of the responses and for only 24 per 
cent of the words. Moreover, word guessability was shown to be less a function 
of using the context than of applying ‘preconceived notions’. And, although 
more proficient students knew more words than less proficient students, they 

---

222 An Introduction to Applied Linguistics
were not any more effective in the use of context. Haynes (1984) also showed 
that students make greater use of local, rather than global, contextual clues in 
their contextual guessing of word meanings, and that what may appear to be 
transparent, ‘guessable’ contexts to native speakers are often incomprehensible 
contexts to non-native speakers.
In a study which examined both guessing from context as well as retention, 
Mondria and Wit-de Boer (1991) found that factors such as ‘subject’, ‘verb’ and 
‘function’ contribute to the guessability of a word in a sentence context, and 
that correctly guessing a word did not lead to improved retention as compared 
with guessing a word incorrectly. In fact, retention of correctly guessed words was 
sometimes even worse than it was for incorrectly guessed words. Mondria and 
Wit-de Boer (1991) conclude that factors that are conducive to guessing are not 
conducive to retention. Since these early studies, there has been accumulating 
evidence that guessing words from context is not very accurate nor very effective 
for specific word meaning retention (see also Nassaji, 2003; Nation, 2001). At the 
same time, it is important to recognize ‘guessing word meanings from context’ as 
a crucial strategy for attempting to maintain comprehension (but not specifically 
learn new word meanings) while reading more difficult texts. Gradually, over 
time, guessing word meanings while maintaining comprehension will also lead 
to important vocabulary gains because of multiple exposures to these words (see 
Grabe, 2009; Schmitt, 2008).
Dictionary Use and L2 Reading
The role of dictionaries in both word learning as well as in reading comprehension 
in second language reading has been of much interest. The early study by 
Bensoussan, Sim and Weiss (1984) of relatively proficient first-year university-
level EFL students in Israel found that use of dictionaries during reading had no 
significant effect on multiple-choice comprehension test scores. Hulstijn (1993) 
found that students with high inferencing ability (that is, were able to guess word 
meaning from context) used a dictionary to the same extent as students with low 
inferencing ability, suggesting that some students may use a dictionary when it 
may not be necessary for comprehension. Thus, these two studies together suggest 
that dictionary use during reading may not be facilitative of second language 
reading comprehension, and possibly unnecessary for higher-proficiency students 
and ineffective for lower-proficiency students.
Implications: In order for L2 learners to read well, they must have an adequately 
sized vocabulary and must be able to recognize the words in that vocabulary 
quickly and accurately. Guessing from context and dictionary use can help in 
acquiring this vocabulary over time, but these skills are not automatic. Rather, they 
need to be developed and practised in order to be used effectively in conjunction 
with reading.
Reading Rate
In L1 reading research studies, there is considerable evidence that fluent readers 
read at rates between 200 wpm and 300 wpm for most types of texts (Carver, 1992). 
Moreover, evidence demonstrated that this fluency develops consistently across 

---

223
Reading
age and grade levels. Unfortunately, L2 students typically do not have 12 years 
to develop fluent reading rates, so recommendations are regularly made to use 
speed reading and reading rate activities in L2 reading classes (see Anderson, 2008; 
Nation 2009, for examples). While there are relatively few published studies of the 
relationship between reading rate development or training and foreign or second 
language reading comprehension, some studies have pointed to the benefits of 
fluency and rate training for improving reading fluency and comprehension. 
Anderson (1991) worked with students to increase reading rate during a 14-week 
semester in a university-level intensive ESL programme and measured students’ 
comprehension scores as well. Students in the experimental group significantly 
increased their reading rate (from 161 wpm to 275 wpm), whereas readers 
in the control group showed an insignificant increase (from 160 wpm to 167 
wpm). Although students in the experimental group did not make significant 
comprehension gains, whilst students in the control group did, the good news 
was that these students’ comprehension did not suffer while their reading rate 
increased dramatically. The results from this study suggest that it is indeed possible 
to help students improve their reading rate. Weigle and Jensen (1996), although 
not including a control group, similarly found significant increases in reading rate 
after training.
In an important set of fluency training studies, Taguchi and Gorsuch (Taguchi, 
Takayasu-Maass and Gorsuch, 2004; Gorsuch and Taguchi, 2008) demonstrated 
that students could improve both their reading fluency and their reading 
comprehension through repeated reading activities. Studies in both Japan and 
Vietnam involved students reading read short texts multiple times to gain practice 
in fluent reading. After a period of treatment, students gained in both reading 
fluency (in terms of WPM reading) and reading comprehension.
Implications: Exercises aimed at improving reading rate seem to help L2 learners, in